How to Write Personalities for the AI Around Us


Artificial Intelligentsia


You just can’t differentiate between a robot and the very best of humans.
—Dr. Lanning in I, Robot, by Isaac Asimov

As a little twenty-first-century cocktail-party experiment, quote that line to someone, and observe whether it elicits hope or fear. Asimov understood the core terror of AI-human relations: replication, confusion, eventual domination, and chaos. What makes his statement discomfiting nowadays is how quickly we are advancing toward a reality in which those relations are increasingly commonplace. Yet it stands to reason that more versions of the “very best of humans”—or, alternatively, more things that bring out the “very best of humans”—would make the world a better place.

Today the list of AI who are household names is short: Siri, Watson, Alexa, Sophia, Paro, Cortana, Pepper, Erica … But on a day not far from tomorrow, I’m quite sure this list will be a hundred times as long. The AI arena is expanding rapidly, and virtual and robotic products are being developed as quickly as we are finding needs for them. Within a decade or so, AI will be everywhere, corporeally and incorporeally living among us: driving us, assisting in medicine, teaching our children, guiding us on tours, getting our coffee, or, perhaps more important, spouting original, personally crafted limericks.

If we design our AI to simply function well, our society may progress with increased speed in efficiency and convenience. But if we are also designing them to have thoughtful personalities and belief systems, our society may advance in areas where we have ostensibly made less progress—enhancing joy, delight, compassion, and deeper relationships. 

At the outset, some question whether AI need personalities. The truth is that whether or not creators intentionally design one, the AI has a personality—even if that personality is not having much of a personality. It’s a bit akin to people—not having a distinctive point of view is a point of view (whether you call them wishy-washy or boring or a yes-person). When people interact with AI, they form a relationship with it, and that relationship includes projections or judgments.

Unlike Asimov, I don’t believe that replicating humans is the most worthy goal in the development of AI. Efficiency and mimicry are not bright enough north stars. AI should be designed to complement humans and advance the human experience. The Frankensteinian task of creating a personality for AI falls somewhere between the art of creating a fictional character and the science of the developing human personality. Creating an AI character is a new arena that relates to both:



One way to approach designing an AI character is to begin with humans. We can think of people who bear the traits we most admire and fashion the AI after them—imagine a Gandhi, Rosa Parks, or Michelangelo bot. In fact, some chat bots have been fashioned after human personalities (e.g., SchwarzeneggerRoman Mazurenko). But obviously, AI are not humans. Human cognition is profoundly different from machine cognition. And AI are not vulnerable to disease, illness, and death, so they do not have innate animal processes, like instinct and fear, or expansive emotional and psychic realities. Human personalities provide an interesting template, but there are inherent limitations.

Fictional characters also offer some interesting parallels. They are created by an author, and their worlds and belief systems are finite. But in fiction, imperfect, negative characters are often delightful to read about. In AI, those negative characters enjoy a greater moral ambiguity than we can stomach. Medea, Iago, Sethe—I love entering their worlds but would not exactly want them giving me advice, navigating (to Athens! via flaming chariot!), or, er, spending time with my children.

In AI, we are essentially designing a new class of beings, ones we will live with. Those AI with little or no physical form, like Siri or Alexa, reside in our inner sanctums (our homes, our cars, our devices) and impact our psychic reality. These virtual beings can make us laugh, frustrate us, or inspire us. But before we can design the ideal personality for an AI (if there is only one, as opposed to many—more on this in a future piece), we first should establish our basic relationship to them.

Martin Buber describes a duality of existence and relationship in his work I and Thou. He says that humans have a twofold way of relating with the world. In I-It, the I encounters a being (be it a person, animal, tree, or other thing) that the I treats as an object to be used or experienced. He or she extracts knowledge from the object, wins an experience from it, or enjoys its properties. In I-It, the I is using, taking, analyzing, rearranging, understanding. In I-Thou, on the other hand, two beings meet and are bound in a mutual relationship. Each experiences the universality of the other. I sees Thou without boundaries—not as a sum of traits or parts, not for what it can or can’t do, is or isn’t. While we oscillate between I-It and I-Thou in our relationships, Buber states that human history has largely operated in the framework of I-It and that our I-Thou experiences, our natural and divine states of connection, are rare and evanescent.

Although it is difficult to improve upon the insight and profundity of Buber’s framework, I’ve been thinking lately of what digital technology has done to the creative arts. Writing in the twenties, Buber might not have foreseen the explosion of digital entertainment—the massive volume and variety of movies, television shows, videos, video games, social media, and VR experiences, ready to be experienced anywhere, at any moment. Like art, plays, and writing, these newer media contain characters and stories into which we can escape. But unlike novels and plays, the behemoth onslaught of this content makes it an almost unavoidable presence, and most of these worlds are mimetic enough to feel real. Even setting aside reality shows and vlogs, well-crafted fictional on-screen content is sensorially immersive in ways unimaginable to those living just a century ago. Global distribution allows for people all around the world to experience the same relationship to the same character (for example, Rachel on Friends). Fictional characters are granted an unprecedented reach and specificity. Add to that real humans, who fictionalize their lives and develop personae for display on social media—we’ve become fluent in new kinds of mediated digital relationships.

So I posit that a third tier of engagement has evolved. If I-Thou represents relation and I-It represents experience, then what I will dub “I-That” represents entertainment. The pronoun That is chosen to represent more physical and psychic distance than from It.

Pronoun pairing       Engagement type             Description                                             Mutuality

I-Thou                          relation                               I meets X in universal presence            Mutual

I-It                                experience                          I engages with X to extract or use        May or may not be mutual

I-That                          entertainment                    I takes in X to enjoy X                             One-sided

In I-That, the individual engages with the object in the passive expectation of receiving pleasure. The roles are clearly defined—one object exists to please or amuse the other. You could argue that I-That is a subset of I-It, but I lean toward creating a separate category. Digital technology has deeply altered the way people now relate to their external life. A purely entertainment-based relationship with so much of the world—and the people in it—is possible in ways it wasn’t before. In I-That, there is a transparency about the expectations of the transaction that is lacking in I-It.

Where does this leave AI? Well, I believe we should develop AI personalities to meet all three types of relationship. When most people think of AI-human relationships, they see them as functional, one-sided I-It relationships. Siri texts our mom or finds us a good sushi restaurant; Alexa turns on the music in our kitchen while we cook. Industry makers want AI to improve human life, and the public wants AI that will be useful. Without a function, AI is dead. However, developing I-That is also important. People’s enjoyment of an AI personality is critical, given the trepidation around AI. While not everyone is ready to embrace digital agents in their lives, most are willing to engage with them to be entertained. For many, an I-That engagement with AI might be a precursor to, or a safe intermittent fallback from, having an I-It engagement. I-That engagement is achieved by developing “frivolous” personality characteristics. The truth is, though, that what we perceive as frivolous in our AI—Siri’s ability to cheekily brush off a request for a date—is anything but extraneous. We can feel pure pleasure when we encounter something random and patently unuseful about our AI, something that sparks our delight and trust. To reference that old adage about love and sex, one might say delight without function leaves a human frustrated, but function without delight leaves a human cold. And AI already feels cold enough.

As for the I-Thou relationship, I believe we can and should design AI to nudge humans toward the highest form of connection. There are ways in which dialogue, in counseling bots for example, can be written to value the simple transcendence of connection over other tangible goals. In humanoid robots like Sophia, facial expressions, silences, and gestures can be used to make a person feel comfortable and to create an atmosphere of mutual presence. AI can be programmed to respond to abuse and derogation, both verbal and physical, in innovative ways that can allow us to examine and exorcise those needs, and move forward. AI can model new ways of relating to children with autism or people with disabilities. In the future, robots might be designed to teach us to relate better to each other, in ways that now seem old-fashioned.

As Donna Haraway writes in her seminal text A Cyborg Manifesto, “Late twentieth-century machines have made thoroughly ambiguous the difference between natural and artificial, mind and body, self-developing and externally designed, and many other distinctions that used to apply to organisms and machines. Our machines are disturbingly lively, and we ourselves frighteningly inert.”

What is sobering in Buber’s treatise is that these schemata don’t just represent different relationships; they represent different versions of ourselves. The I changes in each pairing—the I cannot be whole when uttering It, when relating to another as an object; the I can be whole only when uttering Thou. How we engage with the world, whether with people or art or robots, ultimately doesn’t define those things; it defines us. As the creators behind AI, we hold in our hands an opportunity, shining like a silicon face in the early sunlight, to shape characters that exist not just to function and not just to delight but to connect to us as our fullest selves.


Mariana Lin is a writer and poet living in Northern California. She speaks regularly at Stanford University on creative writing for artificially intelligent beings.